In design reviews I have often heard a choice of vocabulary defended with one sentence: it is the official one. The sentence usually ends the discussion. A body with the right standing has published the model, the model has a proper identifier and a version history, and choosing anything else would need explaining. So the official model goes into the design, and everyone moves on.
That reasoning is sound as far as it goes. It answers one question well, and it quietly assumes the answer to a second question that matters more.
Two qualities that travel under one word
When we call a standard authoritative, we usually mean two different things at once.
The first is authority: the institutional standing of the body that publishes the model. A ministry, a standards organisation or a programme office has the mandate to define certain assets, and its name on a vocabulary carries that mandate with it. Authority is conferred, and it exists from the day of publication.
The second is authoritativeness: the evidence that the parties who are supposed to use the model actually do. Who imports it and builds their own models on it, who maps their terms to it, who runs it in production with live data. Authoritativeness is earned, slowly, by the behaviour of the people around the model, and it can be observed.
Most of the time we treat the two as one. A model with authority is assumed to be authoritative, and a model without a recognised publisher is assumed to be marginal. Both assumptions fail often enough to matter.
When they come apart
Sometimes authority arrives without authoritativeness. A vocabulary is published by exactly the right body, through exactly the right process, and then sees little use among the parties it was written for, because it arrived too late, fitted their systems poorly or competed with something they had already built on. Its standing stays intact, and in reviews it keeps winning the argument, while in production the exchanges run on something else.
The opposite happens too, and it is just as instructive. In June 2011, Google, Microsoft and Yahoo launched schema.org, joined by Yandex later that year: a vocabulary created outside the formal standards bodies, by companies with a shared practical interest. Its authority in the classical sense was modest. Its authoritativeness, for the purpose it served, became enormous, because millions of websites built on it. Anyone choosing a vocabulary for that purpose on letterhead alone would have chosen badly.
The point is that neither quality predicts the other reliably. A modeller who reads only authority is reading half the evidence.
Why the confusion persists
Authority is easy to see. It sits in the header of the specification, in the catalogue entry, in the procurement rule that says to prefer official assets. Authoritativeness is scattered across the published models, catalogues and systems of many organisations, and until recently there was no practical way to read it for a defined circle of peers.
Programme reports follow what is easy to see. They count standards adopted and assets published, which records authority, and they rarely show which parties actually build on what. So the confusion is reinforced every reporting period, and modellers learn that citing the official source is the safe choice, whatever their peers are doing.
Questions to ask of a standard you depend on
- Who publishes it, and with what mandate? That is its authority. Write it down.
- Which of the parties you need to exchange data with import it and build on it?
- Which of them map their own terms to it, and which run it in production? That is its authoritativeness, for your circle and your purpose.
- Where the two disagree, which one is your design following? If the answer is “authority”, ask whether your exchanges will.
- When you report on standards, do you show both?
Reading both, deciding with both
Authority keeps an important place. It gives legitimacy, it settles questions of responsibility and it is often a legal requirement. The improvement is simply to read it next to authoritativeness, and to decide with both in view. A model with high authority and strong adoption among your peers is an easy choice. A model with high authority and little adoption is a warning worth discussing before it goes into a design. Keeping those two readings apart, and visible, is one of the ideas at the heart of EVIDENT, the framework we are developing at METALEVELS.
If you would like to read both qualities for the standards your programme depends on, start a conversation with us.